All-optical synthesis of an arbitrary linear transformation using diffractive surfaces
作者机构：Electrical and Computer Engineering DepartmentUniversity of CaliforniaLos AngelesCA 90095USA Bioengineering DepartmentUniversity of CaliforniaLos AngelesCA 90095USA California NanoSystems InstituteUniversity of CaliforniaLos AngelesCA 90095USA
出 版 物：《Light(Science & Applications)》 (光（科学与应用)（英文版）)
年 卷 期：2021年第10卷第10期
页 面：1928-1948页
核心收录：
学科分类：070207[理学-光学] 07[理学] 08[工学] 0803[工学-光学工程] 0702[理学-物理学]
基 金：The authors acknowledge the US Air Force Office of Scientific Research(AFOSR) Materials with Extreme Properties Program funding(FA9550-21-1-0324)
摘 要：Spatially-engineered diffractive surfaces have emerged as a powerful framework to control light-matter interactions for statistical inference and the design of task-specific optical ***,we report the design of diffractive surfaces to all-optically perform arbitrary complex-valued linear transformations between an input(Ni)and output(No),where Ni and No represent the number of pixels at the input and output fields-of-view(FOVs),***,we consider a single diffractive surface and use a matrix pseudoinverse-based method to determine the complex-valued transmission coefficients of the diffractive features/neurons to all-optically perform a desired/target linear *** addition to this data-free design approach,we also consider a deep learning-based design method to optimize the transmission coefficients of diffractive surfaces by using examples of input/output fields corresponding to the target *** compared the all-optical transformation errors and diffraction efficiencies achieved using data-free designs as well as data-driven(deep learning-based)diffractive designs to all-optically perform(i)arbitrarily-chosen complex-valued transformations including unitary,nonunitary,and noninvertible transforms,(ii)2D discrete Fourier transformation,(iii)arbitrary 2D permutation operations,and(iv)high-pass filtered coherent *** analyses reveal that if the total number(N)of spatially-engineered diffractive features/neurons is≥Ni×No,both design methods succeed in all-optical implementation of the target transformation,achieving negligible ***,compared to data-free designs,deep learning-based diffractive designs are found to achieve significantly larger diffraction efficiencies for a given N and their all-optical transformations are more accurate for NNi×*** conclusions are generally applicable to various optical processors that employ spatially-engineered diffractive surfaces.
维普期刊数据库